What are you looking for?

Showing 9 of 268 Results in Monopolization

TechFreedom Search Engine Regulation Event today

Popular Media Today at 12:30 at the Capitol Visitor Center, TechFreedom is hosting a discussion on the regulation of search engines:  “Search Engine Regulation: A Solution in . . .

Today at 12:30 at the Capitol Visitor Center, TechFreedom is hosting a discussion on the regulation of search engines:  “Search Engine Regulation: A Solution in Search of a Problem?”

The basics:

Allegations of “search bias” have led to increased scrutiny of Google, including active investigations in the European Union and Texas, a possible FTC investigation, and sharply-worded inquiries from members of Congress. But what does “search bias” really mean? Does it demand preemptive “search neutrality” regulation, requiring government oversight of how search results are ranked? Is antitrust intervention required to protect competition? Or can market forces deal with these concerns?

A panel of leading thinkers on Internet law will explore these questions at a luncheon hosted by TechFreedom, a new digital policy think tank. The event will take place at the Capitol Visitor Center room SVC-210/212 onTuesday, June 14 from 12:30 to 2:30pm, and include a complimentary lunch. CNET’s Declan McCullagh, a veteran tech policy journalist, will moderate a panel of four legal experts:

More details are here, and the event will be streaming live from that link as well.  If all goes well, it will also be accessible right here:


Live Broadcasting by Ustream

Filed under: administrative, announcements, essential facilities, google, law and economics, monopolization, regulation, technology Tagged: Declan McCullagh, Eric Goldman, frank pasquale, geoffrey manne, google, james grimmelmann, techfreedom, Web search engine

Continue reading
Antitrust & Consumer Protection

Barnett v. Barnett on Antitrust

Popular Media Tom Barnett (Covington & Burling) represents Expedia in, among other things, its efforts to persuade a US antitrust agency to bring a case against Google . . .

Tom Barnett (Covington & Burling) represents Expedia in, among other things, its efforts to persuade a US antitrust agency to bring a case against Google involving the alleged use of its search engine results to harm competition.  In that role, in a recent piece in Bloomberg, Barnett wrote the following things:

  • “The U.S. Justice Department stood up for consumers last month by requiring Google Inc. to submit to significant conditions on its takeover of ITA Software Inc., a company that specializes in organizing airline data.”
  • “According to the department, without the judicially monitored restrictions, Google’s control over this key asset “would have substantially lessened competition among providers of comparative flight search websites in the United States, resulting in reduced choice and less innovation for consumers.”
  • “Now Google also offers services that compete with other sites to provide specialized “vertical” search services in particular segments (such as books, videos, maps and, soon, travel) and information sought by users (such as hotel and restaurant reviews in Google Places).  So Google now has an incentive to use its control over search traffic to steer users to its own services and to foreclose the visibility of competing websites.”
  • “Search Display: Google has led users to expect that the top results it displays are those that its search algorithm indicates are most likely to be relevant to their query. This is why the vast majority of user clicks are on the top three or four results.  Google now steers users to its own pages by inserting links to its services at the top of the search results page, often without disclosing what it has done. If you search for hotels in a particular city, for example, Google frequently inserts links to its Places pages.”
  • “All of these activities by Google warrant serious antitrust scrutiny. … It’s important for consumers that antitrust enforcers thoroughly investigate Google’s activities to ensure that competition and innovation on the Internet remain vibrant. The ITA decision is a great win for consumers; even bigger issues and threats remain.”

The themes are fairly straightforward: (1) Google is a dominant search engine, and its size and share of the search market warrants concern, (2) Google is becoming vertically integrated, which also warrants concern, (3) Google uses its search engine results in manner that harms rivals through actions that “warrant serious antitrust scrutiny,” and (4) Barnett appears to applaud judicial monitoring of Google’s contracts involving one of its “key assets.”   Sigh.

The notion of firms “coming full circle” in antitrust, a la Microsoft’s journey from antitrust defendant to complainant, is nothing new.   Neither is it too surprising or noteworthy when an antitrust lawyer, including very good ones like Barnett, say things when representing a client that are at tension with prior statements made when representing other clients.  By itself, that is not really worth a post.  What I think is interesting here is that the prior statements from Barnett about the appropriate scope of antitrust enforcement generally, and monopolization in the specific, were made as Assistant Attorney General for the Antitrust Division — and thus, I think are more likely to reflect Barnett’s actual views on the law, economics, and competition policy than the statements that appear in Bloomberg.  The comments also expose some shortcomings in the current debate over competition policy and the search market.

But lets get to it.  Here is a list of statements that Barnett made in a variety of contexts while at the Antitrust Division.

  • “Mere size does not demonstrate competitive harm.”  (Section 2 of the Sherman Act Presentation, June 20, 2006)
  • “…if the government is too willing to step in as a regulator, rivals will devote their resources to legal challenges rather than business innovation. This is entirely rational from an individual rival’s perspective: seeking government help to grab a share of your competitor’s profit is likely to be low cost and low risk, whereas innovating on your own is a risky, expensive proposition. But it is entirely irrational as a matter of antitrust policy to encourage such efforts.
    (Interoperability Between Antitrust and Intellectual Property, George Mason University School of Law Symposium, September 13, 2006)
  • “Rather, rivals should be encouraged to innovate on their own – to engage in leapfrog or Schumpeterian competition. New innovation expands the pie for rivals and consumers alike. We would do well to heed Justice Scalia’s observation in Trinko, that creating a legal avenue for such challenges can ‘distort investment’ of both the dominant and the rival firms.” (emphasis added)
    (Interoperability Between Antitrust and Intellectual Property, George Mason University School of Law Symposium, September 13, 2006)
  • “Because a Section 2 violation hurts competitors, they are often the focus of section 2 remedial efforts.  But competitor well-being, in itself, is not the purpose of our antitrust laws.  The Darwinian process of natural selection described by Judge Easterbrook and Professor Schumpeter cannot drive growth and innovation unless tigers and other denizens of the jungle are forced to survive the crucible of competition.”  (Cite).
  • “Implementing a remedy that is too broad runs the risk of distorting markets, impairing competition, and prohibiting perfectly legal and efficient conduct.” (same)
  • “Access remedies also raise efficiency and innovation concerns.  By forcing a firm to share the benefits of its investments and relieving its rivals of the incentive to develop comparable assets of their own, access remedies can reduce the competitive vitality of an industry.” (same)
  • “The extensively discussed problems with behavioral remedies need not be repeated in detail here.  Suffice it to say that agencies and courts lack the resources and expertise to run businesses in an efficient manner. … [R]emedies that require government entities to make business decisions or that require extensive monitoring or other government activity should be avoided wherever possible.”  (Cite).
  • “We need to recognize the incentive created by imposing a duty on a defendant to provide competitors access to its assets.  Such a remedy can undermine the incentive of those other competitors to develop their own assets as well as undermine the incentive for the defendant competitor to develop the assets in the first instance.  If, for example, you compel access to the single bridge across the Missouri River, you might improve competitive options in the short term but harm competition in the longer term by ending up with only one bridge as opposed to two or three.” (same)
  • “There seems to be consensus that we should prohibit unilateral conduct only where it is demonstrated through rigorous economic analysis to harm competition and thereby harm consumer welfare.” (same)

I’ll take Barnett (2006-08) over Barnett (2011) in a technical knockout.  Concerns about administrable antitrust remedies, unintended consequences of those remedies, error costs, helping consumers and restoring competition rather than merely giving a handout to rivals, and maintaining the incentive to compete and innovate are all serious issues in the Section 2 context.  Antitrust scholars from Epstein and Posner to Areeda and Hovenkamp and others have all recognized these issues — as did Barnett when he was at the DOJ (and no doubt still).  I do not fault him for the inconsistency.  But on the merits, the current claims about the role of Section 2 in altering competition in the search engine space, and the applause for judicially monitored business activities, runs afoul of the well grounded views on Section 2 and remedies that Barnett espoused while at the DOJ.

Let me end with one illustration that I think drives the point home.   When one compares Barnett’s column in Bloomberg to his speeches at DOJ, there is one difference that jumps off the page and I think is illustrative of a real problem in the search engine antitrust debate.  Barnett’s focus in the Bloomberg piece, as counsel for Expedia, is largely harm to rivals.  Google is big.  Google has engaged in practices that might harm various Internet businesses.  The focus is not consumers, i.e. the users.  They are mentioned here and there — but in the context of Google’s practices that might “steer” users toward their own sites.  As Barnett (2006-08) well knew, and no doubt continues to know, is that vertical integration and vertical contracts with preferential placement of this sort can well be (and often are) pro-competitive.  This is precisely why Barnett (2006-08) counseled requiring hard proof of harm to consumers before he would recommend much less applaud an antitrust remedy tinkering with the way search business is conducted and running the risk of violating the “do no harm” principle.  By way of contrast, Barnett’s speeches at the DOJ frequently made clear that the notion that the antitrust laws “protection competition, not competitors,” was not just a mantra, but a serious core of sensible Section 2 enforcement.

The focus can and should remain upon consumers rather than rivals.  The economic question is whether, when and if Google uses search results to favor its own content, that conduct is efficient and pro-consumer or can plausibly cause antitrust injury.  Those leaping from “harm to rivals” to harm to consumers should proceed with caution.  Neither economic theory nor empirical evidence indicate that the leap is an easy one.  Quite the contrary, the evidence suggests these arrangements are generally pro-consumer and efficient.  On a case-by-case analysis, the facts might suggest a competitive problem in any given case.

Barnett (2006-08) has got Expedia’s antitrust lawyer dead to rights on this one.  Consumers would be better off if the antitrust agencies took the advice of the former and ignored the latter.

Filed under: antitrust, doj, economics, error costs, essential facilities, exclusionary conduct, federal trade commission, google, monopolization, technology

Continue reading
Antitrust & Consumer Protection

Manne and Wright on Search Neutrality

Popular Media Josh and I have just completed a white paper on search neutrality/search bias and the regulation of search engines.  The paper is this year’s first . . .

Josh and I have just completed a white paper on search neutrality/search bias and the regulation of search engines.  The paper is this year’s first in the ICLE Antitrust & Consumer Protection White Paper Series:

If Search Neutrality Is the Answer, What’s the Question?


Geoffrey A. Manne

(Lewis & Clark Law School and ICLE)


Joshua D. Wright

(George Mason Law School & Department of Economics and ICLE)

In this paper we evaluate both the economic and non-economic costs and benefits of search bias. In Part I we define search bias and search neutrality, terms that have taken on any number of meanings in the literature, and survey recent regulatory concerns surrounding search bias. In Part II we discuss the economics and technology of search. In Part III we evaluate the economic costs and benefits of search bias. We demonstrate that search bias is the product of the competitive process and link the search bias debate to the economic and empirical literature on vertical integration and the generally-efficient and pro-competitive incentives for a vertically integrated firm to discriminate in favor of its own content. Building upon this literature and its application to the search engine market, we conclude that neither an ex ante regulatory restriction on search engine bias nor the imposition of an antitrust duty to deal upon Google would benefit consumers. In Part V we evaluate the frequent claim that search engine bias causes other serious, though less tangible, social and cultural harms. As with the economic case for search neutrality, we find these non-economic justifications for restricting search engine bias unconvincing, and particularly susceptible to the well-known Nirvana Fallacy of comparing imperfect real world institutions with romanticized and unrealistic alternatives

Search bias is not a function of Google’s large share of overall searches. Rather, it is a feature of competition in the search engine market, as evidenced by the fact that its rivals also exercise editorial and algorithmic control over what information is provided to consumers and in what manner. Consumers rightly value competition between search engine providers on this margin; this fact alone suggests caution in regulating search bias at all, much less with an ex ante regulatory schema which defines the margins upon which search providers can compete. The strength of economic theory and evidence demonstrating that regulatory restrictions on vertical integration are costly to consumers, impede innovation, and discourage experimentation in a dynamic marketplace support the conclusion that neither regulation of search bias nor antitrust intervention can be justified on economic terms. Search neutrality advocates touting the non-economic virtues of their proposed regime should bear the burden of demonstrating that they exist beyond the Nirvana Fallacy of comparing an imperfect private actor to a perfect government decision-maker, and further, that any such benefits outweigh the economic costs.



Filed under: announcements, antitrust, economics, error costs, essential facilities, exclusionary conduct, google, law and economics, markets, monopolization, technology, truth on the market Tagged: antitrust, foundem, google, search, search bias, Search Engines, search neutrality, tradecomet, Vertical integration

Continue reading
Antitrust & Consumer Protection

Net neutrality and Trinko

Popular Media Commentators who see Trinko as an impediment to the claim that antitrust law can take care of harmful platform access problems (and thus that prospective rate . . .

Commentators who see Trinko as an impediment to the claim that antitrust law can take care of harmful platform access problems (and thus that prospective rate regulation (i.e., net neutrality) is not necessary), commit an important error in making their claim–and it is a similar error committed by those who advocate for search neutrality regulation, as well.  In both cases, proponents are advocating for a particular remedy to an undemonstrated problem, rather than attempting to assess whether there is really a problem in the first place.  In the net neutrality context, it may be true that Trinko would prevent the application of antitrust laws to mandate neutral access as envisioned by Free Press, et al.  But that is not the same as saying Trinko precludes the application of antitrust laws.  In fact, there is nothing in Trinko that would prevent regulators and courts from assessing the anticompetitive consequences of particular network management decisions undertaken by a dominant network provider.  This is where the concerns do and should lie–not with an aesthetic preference for a particular form of regulation putatively justified as a response to this concern.  Indeed, “net neutrality” as an antitrust remedy, to the extent that it emanates from essential facilities arguments, is and should be precluded by Trinko.

But the Court seems to me to be pretty clear in Trinko that an antitrust case can be made, even against a firm regulated under the Telecommunications Act:

Section 601(b)(1) of the 1996 Act is an antitrust-specific saving clause providing that “nothing in this Act or the amendments made by this Act shall be construed to modify, impair, or supersede the applicability of any of the antitrust laws.”  This bars a finding of implied immunity. As the FCC has put the point, the saving clause preserves those “claims that satisfy established antitrust standards.”

But just as the 1996 Act preserves claims that satisfy existing antitrust standards, it does not create new claims that go beyond existing antitrust standards; that would be equally inconsistent with the saving clause’s mandate that nothing in the Act “modify, impair, or supersede the applicability” of the antitrust laws.

There is no problem assessing run of the mill anticompetitive conduct using “established antitrust standards.”  But that doesn’t mean that a net neutrality remedy can be constructed from such a case, nor does it mean that precisely the same issues that proponents of net neutrality seek to resolve with net neutrality are necessarily cognizable anticompetitive concerns.

For example, as Josh noted the other day, quoting Tom Hazlett, proponents of net neutrality seem to think that it should apply indiscriminately against even firms with no monopoly power (and thus no ability to inflict consumer harm in the traditional antitrust sense).  Trinko (along with a vast quantity of other antitrust precedent) would prevent the application of antitrust laws to reach this conduct–and thus, indeed, antitrust and net neutrality as imagined by its proponents are not coextensive.  I think this is very much to the good.  But, again, nothing in Trinko or elsewhere in the antitrust laws would prohibit an antitrust case against a dominant firm engaged in anticompetitive conduct just because it was also regulated by the FCC.

Critics point to language like this in Trinko to support their contrary claim:

One factor of particular importance is the existence of a regulatory structure designed to deter and remedy anticompetitive harm. Where such a structure exists, the additional benefit to competition provided by antitrust enforcement will tend to be small, and it will be less plausible that the antitrust laws contemplate such additional scrutiny.

But I don’t think that helps them at all.  What the Court is saying is not that one regulatory scheme precludes the other, but rather that if a regulatory scheme mandates conduct that makes the actuality of anticompetitive harm less likely, then the application of necessarily-imperfect antitrust law is likely to do more harm than good.  Thus the Court notes that

The regulatory framework that exists in this case demonstrates how, in certain circumstances, “regulation significantly diminishes the likelihood of major antitrust harm.”

But this does not say that regulation precludes the application of antitrust law.  Nor does it preclude the possibility that antitrust harm can still exist; nor does it suggest that any given regulatory regime reduces the likelihood of any given anticompetitive harm–and if net neutrality proponents could show that the regulatory regime did not in fact diminish the likelihood of antitrust harm, nothing in Trinko would suggest that antitrust should not apply.

So let’s get out there and repeal that FCC net neutrality order and let antitrust deal with any problems that might arise.

Filed under: antitrust, essential facilities, exclusionary conduct, monopolization, net neutrality, technology Tagged: Competition law, FCC, Federal Communications Commission, Free Press, Monopoly, net neutrality, Network neutrality, regulation, Telecommunications Act

Continue reading
Antitrust & Consumer Protection

Google, Antitrust, and First Principles

Popular Media I’ve read with interest over the last few days the commentary on Microsoft’s filing of a formal complaint with the EU, Microsoft’s defense of its . . .

I’ve read with interest over the last few days the commentary on Microsoft’s filing of a formal complaint with the EU, Microsoft’s defense of its actions, and the various stories around the web.  Geoff and Paul appropriately focus on the error-cost concerns associated with intervention in high-tech markets; Paul also emphasizes the ironies associated with Microsoft’s new status as antitrust complainant.  There will be ample time to discuss the substantive merits of these complaints as they develop.  But I wanted to make two points in this post that I think haven’t received much attention.

The first combines Geoff and Paul’s concerns.  I do not find it particularly interesting that Microsoft has taken this role.  In fact, these developments have been a long time coming.  Competitors frequently attempt to induce regulators to intervene against rivals.  The fact that Microsoft is the competitor certainly is ironic, but it is also expected.  The real lesson about Microsoft’s involvement in this suit is what inferences we should make about antitrust policy in high-tech markets given Microsoft’s quick decline from the alleged dominant and entrenched monopolist that was the focus of the DOJ enforcement action to its current status as rival seeking to employ the antitrust agencies to weaken its rival.  How quickly things change.  Indeed, that is precisely the point.  Dynamic competition, innovation, and quick moving industries do not mean that antitrust rules should never apply; however, these conditions should warrant caution in intervening in the name of consumers without firm evidence of consumer harm.

The second point is that the discussions I’ve read recently focus almost entirely on the question of whether Google favors its own content over that of rivals.  This is not the first time this type of argument has been made in antitrust.  Indeed, the argument is commonly made that whether in search or otherwise (e.g. vertical integration with YouTube), so-called “bias” or “discrimination” in favor of one’s own content is prima facie evidence of a competitive problem that will lead to consumer harm.  That is not so.  Neither economic theory nor empirical evidence support the assertion that, without more, vertical integration is a competitive.  Quite the contrary, the evidence suggests these arrangements are generally pro-consumer and efficient.  On a case-by-case analysis, the facts might suggest a competitive problem in any given case.  The facts, of course, matter.

Unfortunately, most of the discussion of “search bias” and other forms of vertical integration at issue here has focused on whether these arrangements are good or bad for Google’s rivals, not consumers.  First principles tell us that the competitive effects are what matters.  And there is indeed a reason to be skeptical of competitive harm here.  While some commentary in recent days has offered principles of antitrust analysis that are flat out wrong (see, e.g. here, the observation that the antitrust laws require “Microsoft, as a competitor, to have equal access”), we can rely on Professor Hovenkamp to re-focus the discussion on first principles:

“You do need to show consumer harm,” said Herbert Hovenkamp, an antitrust expert at the University of Iowa College of Law. “That becomes more difficult with search engines, where it is easy for consumers to switch to another search engine. That is different than in PC operating systems in the Microsoft case, where the technological lock-in was more obvious.”

Indeed.  It is important to note that this is case is in the EU, not the US.  While both jurisdictions emphasize the need to show consumer harm, the EU appears to be significantly more willing than US courts to infer harm to competition from harm to competitors.  I will leave more substantive analysis of the allegations here for another day as more facts develop.  But I suspect that the ultimate outcome here will turn on, in large part, precisely how faithful the EU analysis is to antitrust’s “first”-first principle: intervention requires evidence of consumer harm.

Filed under: antitrust, economics, error costs, exclusionary conduct, google, monopolization, technology

Continue reading
Antitrust & Consumer Protection

Microsoft comes full circle

Popular Media I am disappointed but not surprised to see that my former employer filed an official antitrust complaint against Google in the EU.  The blog post . . .

I am disappointed but not surprised to see that my former employer filed an official antitrust complaint against Google in the EU.  The blog post by Microsoft’s GC, Brad Smith, summarizing its complaint is here.

Most obviously, there is a tragic irony to the most antitrust-beleaguered company ever filing an antitrust complaint against its successful competitor.  Of course the specifics are not identical, but all of the atmospheric and general points that Microsoft itself made in response to the claims against it are applicable here.  It smacks of competitors competing not in the marketplace but in the regulators’ offices.  It promotes a kind of weird protectionism, directing the EU’s enforcement powers against a successful US company . . . at the behest of another US competitor.  Regulators will always be fighting last year’s battles to the great detriment of the industry.  Competition and potential competition abound, even where it may not be obvious (Linux for Microsoft; Facebook for Google, for example).  Etc.  Microsoft was once the world’s most powerful advocate for more sensible, restrained, error-cost-based competition policy.  That it now finds itself on the opposite side of this debate is unfortunate for all of us.

Brad’s blog post is eloquent (as he always is) and forceful.  And he acknowledges the irony.  And of course he may be right on the facts.  Unfortunately we’ll have to resort to a terribly-costly, irretrievably-flawed and error-prone process to find out–not that the process is likely to result in a very reliable answer anyway.  Where I think he is most off base is where he draws–and asks regulators to draw–conclusions about the competitive effects of the actions he describes.  It is certain that Google has another story and will dispute most or all of the facts.  But even without that information we can dispute the conclusions that Google’s actions, if true, are necessarily anticompetitive.  In fact, as Josh and I have detailed at length here and here, these sorts of actions–necessitated by the realities of complex, innovative and vulnerable markets and in many cases undertaken by the largest and the smallest competitors alike–are more likely pro-competitive.  More important, efforts to ferret out the anti-competitive among them will almost certainly harm welfare rather than help it–particularly when competitors are welcomed in to the regulators’ and politicians’ offices in the process.

As I said, disappointing.  It is not inherently inappropriate for Microsoft to resort to this simply because it has been the victim of such unfortunate “competition” in the past, nor is Microsoft obligated or expected to demonstrate intellectual or any other sort of consistency.  But knowing what it does about the irretrievable defects of the process and the inevitable costliness of its consequences, it is disingenuous or naive (the Nirvana fallacy) for it to claim that it is simply engaging in a reliable effort to smooth over a bumpy competitive landscape.  That may be the ideal of antitrust enforcement, but no one knows as well as Microsoft that the reality is far from that ideal.  To claim implicitly that, in this case, things will be different is, as I said, disingenuous.  And likely really costly in the end for all of us.

Filed under: antitrust, business, exclusionary conduct, law and economics, markets, monopolization, politics, regulation, technology Tagged: Brad Smith, Competition law, European Commission, google, microsoft, politics, regulation

Continue reading
Antitrust & Consumer Protection

Search Bias and Antitrust

Popular Media There is an antitrust debate brewing concerning Google and “search bias,” a term used to describe search engine results that preference the content of the . . .

There is an antitrust debate brewing concerning Google and “search bias,” a term used to describe search engine results that preference the content of the search provider.  For example, Google might list Google Maps prominently if one searches “maps” or Microsoft’s Bing might prominently place Microsoft affiliated content or products.

Apparently both antitrust investigations and Congressional hearings are in the works; regulators and commentators appear poised to attempt to impose “search neutrality” through antitrust or other regulatory means to limit or prohibit the ability of search engines (or perhaps just Google) to favor their own content.  At least one proposal goes so far as to advocate a new government agency to regulate search.  Of course, when I read proposals like this, I wonder where Google’s share of the “search market” will be by the time the new agency is built.

As with the net neutrality debate, I understand some of the push for search neutrality involves an intense push to discard traditional economically-grounded antitrust framework.  The logic for this push is simple.  The economic literature on vertical restraints and vertical integration provides no support for ex ante regulation arising out of the concern that a vertically integrating firm will harm competition through favoring its own content and discriminating against rivals.  Economic theory suggests that such arrangements may be anticompetitive in some instances, but also provides a plethora of pro-competitive explanations.  Lafontaine & Slade explain the state of the evidence in their recent survey paper in the Journal of Economic Literature:

We are therefore somewhat surprised at what the weight of the evidence is telling us. It says that, under most circumstances, profit-maximizing vertical-integration decisions are efficient, not just from the firms’ but also from the consumers’ points of view. Although there are isolated studies that contradict this claim, the vast majority support it. Moreover, even in industries that are highly concentrated so that horizontal considerations assume substantial importance, the net effect of vertical integration appears to be positive in many instances. We therefore conclude that, faced with a vertical arrangement, the burden of evidence should be placed on competition authorities to demonstrate that that arrangement is harmful before the practice is attacked. Furthermore, we have found clear evidence that restrictions on vertical integration that are imposed, often by local authorities, on owners of retail networks are usually detrimental to consumers. Given the weight of the evidence, it behooves government agencies to reconsider the validity of such restrictions.

Of course, this does not bless all instances of vertical contracts or integration as pro-competitive.  The antitrust approach appropriately eschews ex ante regulation in favor of a fact-specific rule of reason analysis that requires plaintiffs to demonstrate competitive harm in a particular instance. Again, given the strength of the empirical evidence, it is no surprise that advocates of search neutrality, as net neutrality before it, either do not rely on consumer welfare arguments or are willing to sacrifice consumer welfare for other objectives.

I wish to focus on the antitrust arguments for a moment.  In an interview with the San Francisco Gate, Harvard’s Ben Edelman sketches out an antitrust claim against Google based upon search bias; and to his credit, Edelman provides some evidence in support of his claim.

I’m not convinced.  Edelman’s interpretation of evidence of search bias is detached from antitrust economics.  The evidence is all about identifying whether or not there is bias.  That, however, is not the relevant antitrust inquiry; instead, the question is whether such vertical arrangements, including preferential treatment of one’s own downstream products, are generally procompetitive or anticompetitive.  Examples from other contexts illustrate this point.

Grocery product manufacturers contract for “bias” with supermarkets through slotting contracts and other shelf space payments.  The bulk of economic theory and evidence on these contracts suggest that they are generally efficient and a normal part of the competitive process.   Vertically integrated firms may “bias” their own content in ways that increase output.  Whether bias occurs within the firm (as is the case with Google favoring its own products) or by contract (the shelf space example) should be of no concern for Edelman and those making search bias antitrust arguments.  Economists have known since Coase — and have been reminded by Klein, Alchian, Williamson and others — that firms may achieve by contract anything they could do within the boundaries of the firm.  The point is that, in the economics literature, it is well known that content self-promoting incentives in a vertical relationship can be either efficient or anticompetitive depending on the circumstances of the situation.  The empirical literature suggests that such relationships are mostly pro-competitive and that restrictions upon the abilities of firms to enter them generally reduce consumer welfare.

Edelman is an economist, and so I find it a bit odd that he has framed the “bias” debate without reference to any of this literature.  Instead, his approach appears to be that bias generates harm to rivals and that this harm is a serious antitrust problem.  (Or in other places, that the problem is that Google exhibits bias but its employees may have claimed otherwise at various points; this is also antitrust-irrelevant.)  For example, Edelman writes:

Search bias is a mechanism whereby Google can leverage its dominance in search, in order to achieve dominance in other sectors.  So for example, if Google wants to be dominant in restaurant reviews, Google can adjust search results, so whenever you search for restaurants, you get a Google reviews page, instead of a Chowhound or Yelp page. That’s good for Google, but it might not be in users’ best interests, particularly if the other services have better information, since they’ve specialized in exactly this area and have been doing it for years.

“Leveraging” one’s dominance in search, of course, takes a bit more than bias.  But I was quite curious about Edelman’s evidence and so I went and looked at Edelman and Lockwood.  Here is how they characterize their research question: “Whether search engines’ algorithmic results favor their own services, and if so, which search engines do so most, to what extent, and in what substantive areas.”  Here is how the authors describe what they did to test the hypothesis that Google engages in more search bias than other search engines:

To formalize our analysis, we formed a list of 32 search terms for services commonly provided by search engines, such as “email”, “calendar”, and “maps”. We searched for each term using the top 5 search engines: Google, Yahoo, Bing, Ask, and AOL. We collected this data in August 2010.

We preserved and analyzed the first page of results from each search. Most results came from sources independent of search engines, such as blogs, private web sites, and Wikipedia. However, a significant fraction – 19% – came from pages that were obviously affiliated with one of the five search engines. (For example, we classified results from youtube.com and gmail.com as Google, while Microsoft results included msn.com, hotmail.com, live.com, and Bing.)

Here is the underlying data for all 32 terms; so far, so good.  A small pilot study examining whether and to what extent search engines favor their own content is an interesting project — though, again, I’m not sure it says anything about the antitrust issues.  No surprise: they find some evidence that search engines exhibit some bias in favor of affiliated sites.  You can see all of the evidence at Edelman’s site (again, to his credit).  Interpretations of these results vary dramatically.  Edelman sees a serious problem.  Danny Sullivan begs to differ (“Google only favors itself 19 percent of the time”), and also makes the important point that the study took place before Yahoo searches were powered by Bing.

In their study, Edelman and Lockwood appear at least somewhat aware that bias and vertical integration can be efficient although they do not frame it in those terms.  They concede, for example, that “in principle, a search engine might feature its own services because its users prefer these links.”  To distinguish between these two possibilities, they conceive of the following test:

To test the user preference and bias hypotheses, we use data from two different sources on click-through-rate (CTR) for searches at Google, Yahoo, and Bing. Using CTR data from comScore and another service that (with users’ permission) tracks users’ searches and clicks (a service which prefers not to be listed by name), we analyze the frequency with which users click on search results for selected terms. The data span a four-week period, centered around the time of our automated searches.  In click-through data, the most striking pattern is that the first few search results receive the vast majority of users’ clicks. Across all search engines and search terms, the first result received, on average, 72% of users’ clicks, while the second and third results received 13% and 8% of clicks, respectively.

So far, no surprises.  The first listing generates greater incremental click-through than the second or third listing.  Similarly, the eye-level shelf space generates more sales than less prominent shelf space.  The authors have a difficult time distinguishing user preference from bias:

This concentration of users’ clicks makes it difficult to disprove the user preference hypothesis. For example, as shown in Table 1, Google and Yahoo each list their own maps service as the first result for the query “maps”. Our CTR data indicates that Google Maps receives 86% of user clicks when the search is performed on Google, and Yahoo Maps receives 72% of clicks when the search is performed on Yahoo. One might think that this concentration is evidence of users’ preference for the service affiliated with their search engine. On the other hand, since clicks are usually highly concentrated on the first result, it is possible that users have no such preference, and that they are simply clicking on the first result because it appears first. Moreover, since the advantage conferred by a result’s rank likely differs across different search queries, we do not believe it is appropriate to try to control for ranking in a regression.

The interesting question from a consumer welfare perspective is not what happens to the users without a strong preference for Google Maps or Yahoo Maps.  Users without a strong preference are likely to click-through on whatever service is offered on their search engine of choice.  There is no significant welfare loss from a consumer who is indifferent between Google Maps and Yahoo Maps from choosing one over the other.

The more interesting question is whether users with a strong preference for a non-Google product are foreclosed from access to consumers by search bias.  When Google ranks its Maps above others, but a user with a strong preference for Yahoo Maps finds it listed second, is the user able to find his product of choice?  Probably if it is listed second.  Probably not if it is delisted or something more severe.  Edelman reports some data on this issues:

Nevertheless, there is one CTR pattern that would be highly suggestive of bias. Suppose we see a case in which a search engine ranks its affiliated result highly, yet that result receives fewer clicks than lower results. This would suggest that users strongly prefer the lower result — enough to overcome the effect of the affiliated result’s higher ranking.

Of course this is consistent with bias; however, to repeat the critical point, this bias does not inexorably lead to — or even suggest — an antitrust problem.  Let’s recall the shelf space analogy.  Consider a supermarket where Pepsi is able to gain access to the premium eye-level shelf space but consumers have a strong preference for Coke.  Whether or not the promotional efforts of Pepsi will have an impact on competition depend on whether Coke is able to get access to consumers.  In that case, it may involve reaching down to the second or third shelf.  There might be some incremental search costs involved.  And even if one could show that Coke sales declined dramatically in response to Pepsi’s successful execution of its contractual shelf-space bias strategy, that merely shows harm to rivals rather than harm to competition.  If Coke-loving consumers can access their desired product, Coke isn’t harmed, and there is certainly no competitive risk.

So what do we make of evidence that in the face of search engine bias, click-through data suggest consumers will still pick lower listings?  One inference is that consumers with strong preferences for content other than the biased result nonetheless access their preferred content.  It is difficult to see a competitive problem arising in such an environment.  Edelman anticipates this point somewhat when observes during his interview:

The thing about the effect I’ve just described is you don’t see it very often. Usually the No. 1 link gets twice as many clicks as the second result. So the bias takes some of the clicks that should have gone to the right result. It seems most users are influenced by the positioning.

This fails to justify Edelman’s position.  First off, in a limited sample of terms, its unclear what it means for these reversals not to happen “very often.”  More importantly, so what that the top link gets twice as many clicks as the second link?  The cases where the second link gets the dominant share of clicks-through might well be those where users have a strong preference for the second listed site.  Even if they are not, the antitrust question is whether search bias is efficient or poses a competitive threat.  Most users might be influenced by the positioning because they lack a strong preference or even any preference at all.  That search engines compete for the attention of those consumers, including through search bias, should not be surprising.  But it does not make out a coherent claim of consumer harm.

The ‘compared to what’ question looms large here.  One cannot begin to conceive of answering the search bias problem — if it is a problem at all — from a consumer welfare perspective until they pin down the appropriate counterfactual.  Edelman appears to assume  — when he observes that “bias takes some of the clicks that should have gone to the right result” — that the benchmark “right result” is that which would prevail if listings were correlated perfectly with aggregate consumer preference.   My point here is simple: that comparison is not the one that is relevant to antitrust.  An antitrust inquiry would distinguish harm to competitors from harm to competition; it would focus its inquiry on whether bias impaired the competitive process by foreclosing rivals from access to consumers and not merely whether various listings would be improved but for Google’s bias.  The answer to that question is clearly yes.  The relevant question, however, is whether that bias is efficient.   Evidence that other search engines with much smaller market shares, and certainly without any market power, exhibit similar bias would suggest to most economists that the practice certainly has some efficiency justifications.  Edelman ignores that possibility and by doing so, ignores decades of economic theory and empirical evidence.  This is a serious error, as the overwhelming lesson of that literature is that restrictions on vertical contracting and integration are a serious threat to consumer welfare.

I do not know what answer the appropriate empirical analysis would reveal.  As Geoff and I argue in this paper, however, I suspect a monopolization case against Google on these grounds would face substantial obstacles.  A deeper understanding of the competitive effects of search engine bias is a worthy project.  Edelman should also be applauded for providing some data that is interesting fodder for discussion.  But my sense of the economic arguments and existing data are that they do not provide the support for an antitrust attack against search bias against Google specifically, nor the basis for a consumer-welfare grounded search neutrality regime.

Filed under: advertising, antitrust, armen alchian, business, economics, exclusionary conduct, google, monopolization, net neutrality, technology

Continue reading
Antitrust & Consumer Protection

Gans on Apple and Antitrust

Popular Media Joshua Gans has an interesting post examining potential antitrust issues involving Apple, an issue we’ve discussed here and here.  Gans focuses in on the two . . .

Joshua Gans has an interesting post examining potential antitrust issues involving Apple, an issue we’ve discussed here and here.  Gans focuses in on the two most relevant issues:

There are two aspects that might raise antitrust concern: (i) Apple’s exclusivity-like requirement that no external payment links be permitted in apps and (ii) Apple’s most-favored customer clause preventing discounting on other platforms. Let’s examine each in turn.

In my earlier post, I emphasized that a potential plaintiff would have a difficult time demonstrating that Apple has monopoly power in any relevant market for the purposes of antitrust analysis.  Both exclusivity arrangements and most-favored customer clauses can generate efficiencies and improve consumer outcomes; they pose little threat to competition and consumers in the absence of durable monopoly power.  I suggested that this was the largest obstacle to any antitrust analysis involving Apple’s subscription model:

The most often discussed bar to an antitrust action against Apple is the one many regulators simply assume into existence: Apple must have market power in an antitrust-relevant market.  While Apple’s share of the smartphone market is only 16% or so, its share of the tablet computing market is much larger.  The WSJ, for example, reports that Apple accounts for about three-fourths of tablet computer sales.  I’ve noted before in the smartphone context that this requirement should not be consider a bar to FTC suit, given the availability of Section 5; however, as the WSJ explains, market definition must be a critical issue in any Apple investigation or lawsuit:

Publishers, for example, might claim that Apple dominates the market for consumer tablet computers and that it has allegedly used that commanding position to restrict competition. Apple, in turn, might define the market to include all digital and print media, and counter that any publisher not happy with Apple’s terms is free to still reach its customers through many other print and digital outlets.

One must conduct a proper, empirically-grounded analysis of the relevant data to speak with confidence; however, it suffices to say that I am skeptical that tablet sales would constitute a relevant market.

Gans agrees, also suggesting that lack of monopoly power undercuts any potential antitrust case against Apple.

Exclusivity can be an issue as it might harm other platforms that might want to sell digital subscriptions. If Apple’s exclusivity means that those platforms cannot generate sales, then a monopoly platform may arise or be sustained. But that is the issue here: where is Apple’s monopoly? It is arguable that Apple has a monopoly over tablet devices and has had that monopoly now for almost 11 months since it first released its iPad. But if a publisher decided not to sell subscriptions for iPad users, it would have other options: particularly the options it had prior to April 2010; web based subscriptions and eReader subscriptions, not to mention physical subscriptions that fall outside of Apple’s terms. It would have to be demonstrated that the iPad was one of the few or the only way to access a particular customer class to believe that publishers were excluded by Apple’s terms. In any case, those terms are not strictly exclusionary as they do not prevent other digital subscription sales – even for iPad access. Instead, they at worst, raise the costs of those other sales. To be sure, raising costs can sometimes be an antitrust violation but the degree of market power a firm would have to possess to make that the case has to be proportionate. Right now, that case appears weak.

Most-favored customer clauses arise when the terms of one supply contract impose conditions on other contracts a party might enter into. Apple is effectively preventing discounting elsewhere. If it did not do this, then that discounting would occur and Apple may be unable to generate as much in sales. Worse than that, Apple may do the hard work of signing consumers up for initial subscriptions only to have those same consumers contacted outside of those arrangements with discounts.

But such clauses can have the effect of raising prices in the market and this is what might concern antitrust authorities. For this to be likely to occur here, Apple must have a requisite degree of power (so that publishers are forced to accept those terms) and it must be the case that prices actually rise. It is too early to tell but if Apple is right and iPad consumers really do purchase more, then it is possible that the price elasticity of demand from iPad consumers is relatively high; that is, charge $10 to an iPad consumer and you generate many more sales than $10 charged for other types of consumers. In this environment, it is not obvious that the iPad will lead to higher digital subscription prices.

My only quibble with Gans’ post is that he appears to describe the “monopoly power” requirement as a problem for antitrust, rather than a sensible requirement that protects consumers from overdeterrence.  For example, Gans writes that “antitrust law, as it currently stands, has difficulty in dealing with industries whereby the path is towards monopoly and how to act prospectively about it.”   Gans does not suggest that the antitrust authorities should bring a case against Apple on these grounds.   But he does seem to imply that antitrust would be more effective if it were more willing to reach business conduct that is not harming consumers, may well be providing significant efficiencies currently, but might generate future harm.   I’m not sure this is what he means or if I’m misinterpreting.  If I’m right, I would have characterized things quite differently, perhaps along the lines of “antitrust law is not willing to sacrifice current gains to consumers for the sake of prohibiting practices that are not currently harming competition and the basis for predicting future harm is speculative, at best.”

Potential minor quibble aside, its a very good post and well worth reading.

Filed under: antitrust, economics, MFNs, monopolization, technology

Continue reading
Antitrust & Consumer Protection

An update on the evolving e-book market: Kindle edition (pun intended)

Popular Media [UPDATE:  Josh links to a WSJ article telling us that EU antitrust enforcers raided several (unnamed) e-book publishers as part of an apparent antitrust investigation . . .

[UPDATE:  Josh links to a WSJ article telling us that EU antitrust enforcers raided several (unnamed) e-book publishers as part of an apparent antitrust investigation into the agency model and whether it is “improperly restrictive.”  Whatever that means.  Key grafs:

At issue for antitrust regulators is whether agency models are improperly restrictive. Europe, in particular, has strong anticollusion laws that limit the extent to which companies can agree on the prices consumers will eventually be charged.

Amazon, in particular, has vociferously opposed the agency practice, saying it would like to set prices as it sees fit. Publishers, by contrast, resist the notion of online retailers’ deep discounting.

It is unclear whether the animating question is whether the publishers might have agreed to a particular pricing model, or to particular prices within that model.  As a legal matter that distinction probably doesn’t matter at all; as an economic matter it would seem to be more complicated–to be explored further another day . . . .]

A year ago I wrote about the economics of the e-book publishing market in the context of the dispute between Amazon and some publishers (notably Macmillan) over pricing.  At the time I suggested a few things about how the future might pan out (never a god good idea . . . ):

And that’s really the twist.  Amazon is not ready to be a platform in this business.  The economic conditions are not yet right and it is clearly making a lot of money selling physical books directly to its users.  The Kindle is not ubiquitous and demand for electronic versions of books is not very significant–and thus Amazon does not want to take on the full platform development and distribution risk.  Where seller control over price usually entails a distribution of inventory risk away from suppliers and toward sellers, supplier control over price correspondingly distributes platform development risk toward sellers.  Under the old system Amazon was able to encourage the distribution of the platform (the Kindle) through loss-leader pricing on e-books, ensuring that publishers shared somewhat in the costs of platform distribution (from selling correspondingly fewer physical books) and allowing Amazon to subsidize Kindle sales in a way that helped to encourage consumer familiarity with e-books.  Under the new system it does not have that ability and can only subsidize Kindle use by reducing the price of Kindles–which impedes Amazon from engaging in effective price discrimination for the Kindle, does not tie the subsidy to increased use, and will make widespread distribution of the device more expensive and more risky for Amazon.

This “agency model,” if you recall, is one where, essentially, publishers, rather than Amazon, determine the price for electronic versions of their books sold via Amazon and pay Amazon a percentage.  The problem from Amazon’s point of view, as I mention in the quote above, is that without the ability to control the price of the books it sells, Amazon is limited essentially to fiddling with the price of the reader–the platform–itself in order to encourage more participation on the reader side of the market.  But I surmised (again in the quote above), that fiddling with the price of the platform would be far more blunt and potentially costly than controlling the price of the books themselves, mainly because the latter correlates almost perfectly with usage, and the former does not–and in the end Amazon may end up subsidizing lots of Kindle purchases from which it is then never able to recoup its losses because it accidentally subsidized lots of Kindle purchases by people who had no interest in actually using the devices very much (either because they’re sticking with paper or because Apple has leapfrogged the competition).

It appears, nevertheless, that Amazon has indeed been pursuing this pricing strategy.  According to this post from Kevin Kelly,

In October 2009 John Walkenbach noticed that the price of the Kindle was falling at a consistent rate, lowering almost on a schedule. By June 2010, the rate was so unwavering that he could easily forecast the date at which the Kindle would be free: November 2011.

There’s even a nice graph to go along with it:

So what about the recoupment risk?  Here’s my new theory:  Amazon, having already begun offering free streaming videos for Prime customers, will also begin offering heavily-discounted Kindles and even e-book subsidies–but will also begin rescinding its shipping subsidy and otherwise make the purchase of dead tree books relatively more costly (including by maintaining less inventory–another way to recoup).  It will still face a substantial threat from competing platforms like the iPad but Amazon is at least in a position to affect a good deal of consumer demand for Kindle’s dead tree competitors.

For a take on what’s at stake (here relating to newspapers rather than books, but I’m sure the dynamic is similar), this tidbit linked from one of the comments to Kevin Kelly’s post is eye-opening:

If newspapers switched over to being all online, the cost base would be instantly and permanently transformed. The OECD report puts the cost of printing a typical paper at 28 per cent and the cost of sales and distribution at 24 per cent: so the physical being of the paper absorbs 52 per cent of all costs. (Administration costs another 8 per cent and advertising another 16.) That figure may well be conservative. A persuasive looking analysis in the Business Insider put the cost of printing and distributing the New York Times at $644 million, and then added this: ‘a source with knowledge of the real numbers tells us we’re so low in our estimate of the Times’s printing costs that we’re not even in the ballpark.’ Taking the lower figure, that means that New York Times, if it stopped printing a physical edition of the paper, could afford to give every subscriber a free Kindle. Not the bog-standard Kindle, but the one with free global data access. And not just one Kindle, but four Kindles. And not just once, but every year. And that’s using the low estimate for the costs of printing.

Filed under: antitrust, business, cartels, contracts, doj, e-books, economics, error costs, law and economics, litigation, MFNs, monopolization, resale price maintenance, technology, vertical restraints Tagged: agency model, Amazon, Amazon Kindle, antitrust, Apple, doj, e-book, e-books, iBookstore, Kindle, major publishers, MFN, most favored nations clause, per se, price-fixing, publishing industry, Rule of reason, two-sided markets, vertical restraints

Continue reading
Antitrust & Consumer Protection